Anthropic's 80,000-Person Survey Reveals a New Landscape of Public AI Perception
Anthropic used its own AI models to conduct in-depth interviews with approximately 80,000 users, finding that public att…
📂 Research
Latest articles in Research
The Era of AI Training AI Has Arrived: An In-Depth Look at Three Cutting-Edge Trends
ImportAI Issue 449 reveals three major trends: large models training each other as a new paradigm, 72-billion-parameter …
AI Kernels, Decentralized Training, and Universal Representations: Three Frontier Trends
Import AI Issue 439 spotlights three cutting-edge directions: high-efficiency AI compute kernels, decentralized distribu…
AI Weekly: LLM Societies, Huawei's AI-Powered Chip Development, and Quantifying Creativity
Import AI Issue 444 spotlights three cutting-edge topics: large language models building virtual societies, Huawei lever…
2026 May Become a Critical Turning Point for Superintelligence
Import AI Issue 445 highlights three frontier developments: AI successfully tackling cutting-edge mathematical proofs, t…
Import AI 450: China's Electronic Warfare Model and New Challenges in LLM Security
This edition of Import AI focuses on three cutting-edge topics: China's release of an electronic warfare AI model, resea…
Automation of AI Alignment Research and Safety Evaluation of Chinese Models Draw Industry Attention
Import AI Issue 454 focuses on three cutting-edge topics: the automation of alignment research, safety evaluation of Chi…
UK Biobank Data Incident Sounds Alarm on AI Medical Data Security
UK Biobank head Professor Rory Collins described the data misuse incident as caused by 'a few bad apples,' expressing an…
New Study Finds Plants Can 'Hear' Rain, With AI Helping Unlock the Mystery
A new study reveals that plants can perceive the sound of falling rain and produce physiological responses. Using AI aco…
Bionic Memory Decay AI System Debuts with 52% Recall Rate Simulating Human Forgetting
A developer showcased an AI memory system on Hacker News that mimics biological memory decay mechanisms, intentionally r…
QIMMA: The First Quality-First Leaderboard for Arabic Large Language Models
Arabic AI evaluation reaches a major milestone. The QIMMA leaderboard, built around a 'quality-first' philosophy, provid…
Ecom-RLVE: A New Adaptive Verifiable Reinforcement Learning Paradigm for E-Commerce Conversational AI
A research team has proposed the Ecom-RLVE framework, which leverages reinforcement learning to optimize decision-making…